Re: [Boost-users] [asio] Async (or sync) in threads

We're good. No worries... On Thu, Jul 25, 2013 at 3:22 PM, Anthony Foiani wrote:

Michael Powell writes:

...

I find it hard to imagine that anyone is using Boost.Asio, at least targeting Linux flavors and/or for ARM.

Dunno about ARM, but I've been using Boost.ASIO (since version 1.43 or so, currently on 1.51) on Linux (x86-64 and ppc32, kernel versions 2.6.2x through 3.9.x) with great success.

Doesn't mean that there's not potential issues -- ARM is a very fragmented "architecture" to support, with a huge number of variants and licensed cores etc. So it's possible that you're doing everything right, but something in your toolchain is failing you. (That "illegal instruction" is worrisome, and makes me wonder if your toolchain is 100% aligned with your hardware.)

You aren't kidding re: ARM is a very fragmented architecture to support. As you can imagine, "toolchain" is usually among the foremost inter- and intra-office discussions. More than it needs to be IMHO. It's possible that it's not. I just looked and I believe we were possibly (stronger, probably) not targeting the correct processor.

...

Anyone? Thank you...

You might want to start debugging this by trying to get the ASIO tutorials and examples running on your target hardware.

If time permits I shall. For now I am going with a "simpler" client/server socket implementation. It's a little closer to the socket itself, for what we need to get done.

If that works, then we can see whether you're using ASIO in a way that is unexpected.

Willing to admit, possibly not. I don't think it's that complicated. A certain amount of faith I am putting in the io_service scheduling its call backs, etc.

Good luck, Tony _______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users